Retinal Vessels Change in PrimaryAngle-Closure Glaucoma: The HandanEye StudyJianlu Gao1,2, Yuanbo Liang1, Fenghua Wang1, Ran Shen3, Tienyin Wong4,5, Yi Peng1,David S. Friedman6,7 & Ningli Wang1

1Beijing Tongren Eye Center, Capital University ofMedical Sciences, Beijing, China, 2Liaocheng Clinical Hospital, TaishanMedicalCollege, Shandong Province, China, 3Handan Eye Hospital, Hebei Province, People’s Republic of China, 4Centre for Eye ResearchAustralia, University of Melbourne, Melbourne, Australia, 5Singapore Eye Research Institute, National University of Singapore,Singapore, Republic of Singapore, 6Dana Center for Preventive Ophthalmology, Wilmer Eye Institute, Johns Hopkins University,Baltimore, Maryland, 7Department of International Health, Bloomberg School of Public Health, Johns Hopkins University, Baltimore,Maryland.

Glaucoma is the second leading cause of blindness worldwide. To examine the relationship between angleclosure and the retinal vessel diameter in Chinese adults, we conducted Handan Eye Study (HES), a largepopulation-based cross-sectional study, which enrolled 6830 participants .30 year-old living in 13randomly selected villages of Yongnian County. After adjusting for age, gender, spherical equivalent (SE),diabetes, and hypertension, the mean central retinal artery equivalent (CRAE, mm) was 127.1 6 7.0 and145.6 6 4.4 in primary open-angle glaucoma (POAG) and primary angle closure glaucoma (PACG),respectively; narrower than that in normal control (156.1 6 0.4), primary angle-closure suspect (PACS)(156.3 6 1.1) or primary angle closure (PAC) (156.0 6 3.4) (P 5 0.001). The mean central retinal veinequivalent (CRVE, mm) was 229.06 5.9 and 215.86 9.5 in POAG and PACG, respectively; narrower thanthat in normal control (238.36 0.5), PACS (241.26 1.4) or PAC (242.26 4.6) (P5 0.001). There was nosignificant difference in the mean CRAE or CRVE between PACG and POAG. Compared to the normalcontrol (0.66), the mean arterio-venous ratio (AVR) was smaller in POAG (0.64) and PACG (0.59), whereaslarger in PACS (0.65) and PAC (0.67) (P 5 0.003). To conclude, PACG and POAG individuals havenarrower retinal arteries and veins.

G laucoma is a critical public health problem and the second (after cataracts) leading cause of blindnessworldwide. Primary open-angle glaucoma (POAG) is themajor reason for incurable visual impairment1.It is characterized by progressive loss of retinal ganglion cells, resulting in glaucomatous optic neuro-

pathy (GON). The etiology of POAG is not well established. Potential risk factors include intraocular pressure(IOP), age2, race, family history, genetic3, myopia4, intraocular pressure5, and poor perfusion of optic nerve heador ganglion cell layer6–9. The common consequence for potential etiologies is optic nerve head damage, which issecondary to primarily ganglion cell axon loss; however, loss of blood vessels and glial cells has also been observed.There aremany postulatedmechanisms of ganglion cell damage. It has been reported that eyes with glaucomatousdamage had narrower retinal arteriolar diameters (183 6 2.6 mm) than normal eyes (194 6 0.4 mm) or ocularhypertension (195 6 1.6 mm)8. A prospective cross-sectional study observed significant reduction in the speedand flow of retinal blood in POAG patients compared to normal controls. It is unclear whether an ischemicprocess can lead to GON or whether the narrowing of retinal blood vessel is secondary to glaucoma10.

In primary angle closure (PAC), an eye has a primary anatomic narrow angle and trabecular obstruction in theperipheral iris, caused by peripheral anterior synechiae (PAS), elevated IOP, iris whorling or sectoral atrophy, andexcessive pigment deposition on the trabecular surface. The eye does not have glaucomatous damage of the opticnerve.

In primary angle-closure suspect (PACS) or anatomic narrow angle, the anterior chamber angle recess has anabnormally narrow angular width. The peripheral iris is located close to, but not touching, the posterior pig-mented trabecular meshwork (TM). No PAS are present. IOP, optic nerve, and visual field are normal.

Primary angle closure glaucoma (PACG) is diagnosed when iridotrabecular contact is present in three or morequadrants of the drainage angle in the presence of documented optic nerve damage and visual field loss. It has

OPEN

SUBJECT AREAS:EYE ABNORMALITIES

GLAUCOMA

Received2 December 2014

Accepted2 March 2015

Published

Correspondence andrequests for materials

should be addressed toN.W. (wningli@vip.

163.com)

SCIENTIFIC REPORTS | 5 : 9585 | DOI: 10.1038/srep09585 1

30 April 2015

been proposed that the GON in PACG is caused by high IOP andreduced circulation5,11. We hypothesize that the retinal vessel dia-meter of PACG patients may confer the changes in circulation andcorrelate with the subsequent GON. In China, 1.0–1.5% of people$40 year-old have PACG6–9, which is much higher than that inCaucasian (0.1%–0.6%)12,13. This high prevalence has enabled theHandan Eye Study (HES), a large population-based study on eyediseases in northern China, obtained a larger group of PACGpatients. Here, we analyzed the differences in retinal vessel diametersamong PACS, PAC, PACG, POAG and normal participants, afteradjusting for age, gender, refraction, and other possible confounders.

MethodsStudy population and design. The HES is a population-based study on vision,common eye diseases, and other outcomes of health problems. The study wasconducted according to the Declaration of Helsinki, and was approved by EthicalCommittee of Beijing Tongren Hospital. Written informed consent was obtainedfrom each of the participants. Details of the study design, sampling plan, and baselinedata have been described previously14,15. In brief, 7557 eligible individuals older than30 years living in Yongnian County, Handan City, Hebei Province, China, wereidentified from 13 randomly selected villages using a stratified, clustered, and multi-staged sampling technique, with probabilities proportionate to the size of thepopulation in each cluster. Among the population in Yongnian County, 90% arefarmers, and 98% are Han Chinese.

All eligible subjects were invited to visit Yongnian County Hospital for a serialexamination. Those who declined to visit the hospital were offered a simplifiedevaluation at a temporary field site established in the village; those who declined tovisit the temporary site were offered a limited examination conducted at home. Allfieldwork was conducted from October 2006 to October 2007. Finally, 6830 subjectscompleted the examination, and 6656 of 6830 (97.45%) participates with detaileddemographic and clinical information were included in final data analysis.

Procedures and definitions. For the baseline evaluation, 5909 of 6830 participantswere examined in the county hospital, 807 in a temporary study site at the village, and114 at home. At the county hospital and the village temporary site, fundusphotographs of two fields (optic disc and macula) were taken from both eyes of eachparticipant after pupil dilation using a digital retinal camera. The photographs weretaken by Canon CR-DGi with a 20D SLR back (Canon, Japan). Intraocular pressure(IOP)wasmeasuredwith a Kowa applanation tonometerHA-2 (KowaCompany Ltd.Tokyo, Japan). Visual field test was conducted using the Humphrey Visual FieldAnalyzer 750i (Carl Zeiss, Jena, Germany) in all participants who were suspected tohave glaucoma. Lens thickness and axial measurements were obtained using an A-mode ultrasound device. Other details of the eye examinations have been describedpreviously, including the baseline visual acuity (VA), and the spherical equivalent(SE) measured using subjective refraction7,14.

The retinal vessel diameter measurement was performed as previouslydescribed8,16. Graders were trained at the Retinal Vascular Imaging Centre at TheUniversity of Melbourne prior to the assessment of the photographs. Retinal vesseldiameters from digital retinal images were measured using a computer-based pro-gram by trained graders who were blinded to the status of each of the participant. Foran image, all arterioles and venules coursing through an area of one-half to one-discdiameter away from the optic disc margin were measured and summarized as thecentral retinal artery equivalent (CRAE) and central retinal vein equivalent (CRVE),respectively. These equivalents represent the average caliber of arterioles and venulesof the eye (Figure 1).

The limbal anterior chamber depth (LACD) was measured by slit lamp examina-tion. Gonioscopywas performed in one out of every ten participants as well as in thosewith LACD #40% corneal thickness using Goldmann Magnaview lens (OcularInstruments, Bellevue, Washington, USA) at 253 magnification under low ambientillumination by an experienced ophthalmologist. A narrow vertical beam of 1 mm inlength was offset vertically for superior and inferior quadrants, and horizontally for

nasal and temporal quadrants. Small movements of the lens were allowed to visualizethe drainage angle, but large movements were avoided because of the possibility ofindentation. Dynamic examinations with Goldmann lens were performed after thestatic gonioscopy of the four quadrants was completed. If a satisfactory examinationcould not be achieved with Goldmann lens, 4-mirror Sussman lens (OcularInstruments, Bellevue, Washington, USA) were used. The Spaeth GonioscopicGrading System was used to record the results17.

The vertical cup-to-disc ratio (VCDR) was calculated to exclude the presence ofperipapillary atrophy and the scleral ring of Elschnig. The margin of the cup wasdefined by stereoscopic view as the point of the maximum inflection of the vesselscrossing the neuroretinal rim. Standard photographs for VCDR from 0.1 to 1.0 in 0.1increment were used for the grading process.

Every 10th participate was systematically sampled with a visual field test using 24-2Swedish Interactive Threshold Algorithm (SITA) fast program. In addition, all par-ticipants with angle closure disease or suspected glaucoma had SITA standard visualfield test. Tests were repeated twenty minutes later if the glaucoma hemifield test(GHT) was outside the normal limits, on the borderline, or if the test was unreliable (i.e., fixation loses .20%, false positives .33%, or false negatives .33%).

GON was diagnosed based on the consensus opinion of three glaucoma specialistsafter review of all relevant information including patient history, VA, and optic nervestereo photographs. POAGwas diagnosed if 1) GONwas present without identifiablesecondary causes; and 2) angle closure was excluded on gonioscopy18. PACS wasdiagnosed if the posterior trabecular meshwork was not visible for 180u or more ongonioscopy. PAC was defined as PACS with IOP $21 mmHg or the presence ofperipheral anterior synechiae (PAS) without GON. PACG was defined as PAC withGON13. The eyes without PACS, PAC, PACG, POAG, or non-glaucomatousneuropathy were defined as the normal.

Certified nurses measured height, weight, pulse rate, and blood pressure accordingto standardized protocols. Total cholesterol, total triglycerides, low density lipopro-tein (LDL) cholesterol, high density lipoprotein (HDL) cholesterol, and glucose levelin fasting blood were analyzed using certified protocols. Diabetes mellitus (DM) andhypertension were defined as previously reported14. Migraine and angina werediagnosed using interviewer-administrated questionnaires.

Statistics. Patients diagnosed with ocular hypertension without GON were assignedto the ‘‘normal’’ group(control), and patients diagnosed with normal-tensionglaucoma were assigned to the POAG group, because previous studies have found nodifference in retinal vessel diameter among these groups8.

All statistical analysis was performed using SAS software9,13. Data from the righteyes of all participants in the normal controls and binocular glaucoma participantswas analyzed. Data from the glaucomatous eye in participants with monocularglaucoma was analyzed. The baseline data was analyzed using covariance, adjusted bypotentially confounding variables including age, gender, intraocular pressure (IOP),spherical equivalent refractive error (SE), diabetes (DM), and hypertension in logisticregression models. ANCOVA was used to adjust the influences of age, gender, IOP,SE, DM, and hypertension on retinal vessel diameter or its measurement. A P value,0.05 was considered as statistically significant.

ResultsA total of 6716 participants completed the ocular examination. Sixtyparticipants were excluded because the fundus images were non-gradable. A total of 6656 participants (97.45%) were included in dataanalysis(3088male,3568female). Angle closure was diagnosed if theposterior pigmented trabecular meshwork was invisible for $180u.Primary angle-closure suspect (PACS) was diagnosed if iris tra-becular contact $180 degree with intraocular pressure (IOP),21 mmHg, without peripheral anterior synechiae nor glaucomat-ous optic nerve damage. Primary angle closure (PAC) was diagnosedif angle closure presented with IOP $21 mmHg and/or peripheralanterior synechiae, without glaucomatous optic neuropathy. Primaryangle closure glaucoma (PACG) was diagnosed if angle closure and

Figure 1 | The fundus photographs of Control (A), PACG (B) and POAG (C), and the measured area (between2 blue circles 0.5 to 1.0-disc diameteraway from the optic disc margin) for retinal arteriole and vein diameter.

www.nature.com/scientificreports

SCIENTIFIC REPORTS | 5 : 9585 | DOI: 10.1038/srep09585 2

glaucomatous nerve damage presented. Average retinal vessel dia-meters were summarized as arteriolar and venular equivalents.Among them, 731 patients were diagnosed as PACS(), 64 as PAC,19 as PACG, and 54 as POAG. The baseline characteristics of all ofthe participants were shown in Table 1. Participants with GON wereolder and had a higher prevalence of DM and hypertension. The IOPwas increased in PACG, PAC and POAG groups. The VCDR inPACG (0.85) and POAG (0.66) groups were larger than that inthe normal control (0.41), PACS (0.42) and PAC (0.40) groups.The participants with PACS and PAC were more likely to havehyperopia, and those with POAG had a higher prevalence of myopicand hyperlipidemia. After adjusting for age and gender, there were

significant differences in SE, IOP, DM, and hypertension amongdifferent groups (Figure 2).Figure 3 summarized the retinal arteriole diameters in the normal

and glaucomatous groups after adjusting for age, gender, SE, DM,and hypertension. Themean CRAEwas significantly narrower in thePACG (127.1 6 7.0 mm) and POAG (145.5 6 4.4 mm) groups thanthat in the normal control (156.16 0.4 mm), PACS (156.36 1.1 mm),and PAC (156.06 3.4 mm) groups (P 5 0.001). However, the extentof narrowing in CRAE was comparable between PACG and POAGgroup (95% CI: 21.5 mm to 35.6 mm).As shown in Figure 4, the mean CRVE was significantly narrower

in the PACG (215.8 6 9.5 mm) and POAG (229.0 6 5.9 mm) than

Table | Characteristics of participants for retinal vessel examination in the Handan Eye study

Normal (n 5 5788) PACS (n 5 731) PAC (n 5 64) POAG (n 5 54) PACG (n 5 19) P*

Continuous variables (mean 6 SD)Age (years) 50.8 6 11.8 59.7 6 9.0 61.3 6 8.9 61.2 6 10.4 64.0 6 13.9 ,0.001BMI 24.5 6 3.8 24.4 6 3.7 25.2 6 4.0 23.6 6 3.1 23.9 6 4.3 0.276SE (D) 20.22 6 1.79 0.64 6 1.28 0.52 6 2.14 20.68 6 2.59 0.08 6 1.48 ,0.001IOP (mm Hg) 15.0 6 2.8 14.8 6 2.8 19.0 6 5.0 16.9 6 3.11 24.3 6 13.4 ,0.001Pulse (times/min) 77.4 6 2.0 78.9 6 12.2 81.2 6 14.0 80.2 6 13.4 77.4 6 13.9 0.002MBP (mm Hg) 97.4 6 14.2 100.4 6 14.0 103.4 6 13.7 102.5 6 16.3 100.3 6 13.1 ,0.001CRAE (mm) 154.9 6 26.6 155.4 6 20.2 159.0 6 38.7 143.7 6 20.5 128.6 6 35.8 ,0.001CRVE (mm) 235.9 6 36.2 240.7 6 30.0 241.1 6 41.0 223.9 6 33.6 231.3 6 57.4 ,0.001AVR 0.66 6 0.09 0.65 6 0.09 0.68 6 0.23 0.65 6 0.11 0.59 6 0.22 ,0.001VCDR 0.41 6 0.11 0.42 6 0.10 0.40 6 0.10 0.66 6 0.18 0.85 6 0.18 ,0.001

Categorical variables (%)Migraine 6.05 7.57 9.52 6.82 11.11 0.214Hypertension 45.13 63.20 70.31 73.33 73.68 ,0.001DM 6.14 10.29 12.50 14.29 0.00 ,0.001Hyperlipidemia 36.16 37.08 50.91 56.76 29.41 0.016

Normal eye: eyes of participants without glaucoma (PACS, PAC, PACG, POAG) or non-glaucomatous neuropathy; BMI: bodymass index; SE (D): spherical equivalent (diameter); IOP: intraocular pressure;MBP: mean arterial blood pressure; DM: diabetes mellitus; CRAE: central retinal artery equivalent; CRVE: central retinal vein equivalent; AVR: ratio of CRAE/CRVE. VCDR: vertical cup disc ration. VCDR:vertical cup-to-disc ratio. *P value was from chi-square test (2-sided).

Figure 2 | Percent of Migraine, Hypertension, DM, and Hyperlipidemia in Control(Non-glaucoma)/PACs/PAC/POAG/PACG group.

www.nature.com/scientificreports

SCIENTIFIC REPORTS | 5 : 9585 | DOI: 10.1038/srep09585 3

that in the normal control (238.3 6 0.5 mm), PACS (241.2 61.4 mm), and PAC (242.2 6 4.6 mm) (P 5 0.012). However, theextent of narrowing in CRVE was comparable between PACG andPOAG (95% CI: 214.2 mm to 35.8 mm).The mean AVR was shown in Figure 5. After adjusting for age,

gender, SE, DM, and hypertension, the AVR in PACG (0.59) groupwas smaller than that in the normal control (0.66) (P5 0.003). Therewas no significant change in the AVR from PACS (0.65), PAC (0.67)and POAG (0.64) groups compared to the normal control.

DiscussionIn this HES, retinal blood vessel diameter was measured by themethod used in Blue Mountains Eye Study (BMES) and BeaverDam Eye Study (BDES)8,16. The influence of age, gender, IOP, SE,DM, and hypertension on the vessel diameter or its measurementwas adjusted by ANCOVA. We found that the diameters of retinalartery and retinal vein in POAG were significantly narrower thanthose in the normal control (P5 0.001), with a reduction of 10.6 mm(6.8%) and 9.3 mm (3.9%), respectively. This result was consistentwith those reported by BMES and BDES8,16, and another Chinesepopulation-based study named Beijing Eye Study (BES), where theretinal artery was narrower in the glaucomatous eyes19. The BES didnot use the same method to measure retinal vessel diameter and didnot separately analyze the data from the PACG participates. In ourHES population, 19 participants with PACG, 64 with PAC and 731with PACS had data of retinal vessel diameters. The relatively morePACG individuals in our study may result from the large sample sizeof HES and higher prevalence of PACG (as high as 1.0% amongpeople aged $40 year-old) in Chinese population20. After adjusted

the influence of age, gender, IOP, SE, DM, and hypertension, thediameters of the retinal artery and retinal vein in PACGwere smallerthan those in normal control, with a reduction of 29.0 mm (18.6%)and 22.5 mm (9.4%), respectively. So our data showed that retinalvessel narrowing exists in not only POAG, but also in other kind ofglaucomatous eyes such as PACG. To our knowledge, very few stud-ies on retinal vessel diameter in PACG had been reported beforeHES, especially such a large population-based study. However, ourdata did not provide evidence that angle closure process is related tothe change of retinal vessel diameter.Previous hospital-based and population-based studies conducted

in Europeans and Africans reported the narrowing of peripapillaryretinal blood vessels in patients with glaucoma8,21,22. The blood cir-culation contributes to the development and progression of GON inPOAG patients. However, the Rotterdam Study, which was a popu-lation-based prospective cohort study23, analyzed the baseline retinalvessel diameters of 5517 participants and found no significant dif-ference between non-glaucoma participants and 74 POAG indivi-duals after a mean follow-up time of 6.5 years. Their results mayindicate that blood vessel narrowing in glaucoma is part of the dis-ease process instead of a cause of optic nerve damage. In ourHES, theCRAE and CRVE were narrower in PACG and POAG than those innormal control, PAC or PACS. This finding further supports thenotion that the narrowing of retinal vessels results from the glaucomaprocess.The elevated IOP and structure changes of optic cup in eyes of

glaucoma were considered to be related to the retinal vein obstruc-tion. A retrospective study found that 18 of 83 patients initially pre-senting with central retinal vein occlusion had been diagnosed with

Figure 3 | The retinal arteriole diameters (CRVE) compared to the Control(Non-glaucoma) group, *p , 0.05.

www.nature.com/scientificreports

SCIENTIFIC REPORTS | 5 : 9585 | DOI: 10.1038/srep09585 4

Figure 4 | The retinal vein diameters (CRVE) compared to the Control(Non-glaucoma) group, *p , 0.05.

Figure 5 | The retinal arteriole/vein diameter (CRAE/CRVE) and arteriole/vein ratio (AVR) in Control(Non-glaucoma) and glaucoma groups. P 5

0.003 compared between PACG and Control(Non-glaucoma) groups.

www.nature.com/scientificreports

SCIENTIFIC REPORTS | 5 : 9585 | DOI: 10.1038/srep09585 5

POAG25. A hospital-based study found that central retinal vein col-lapse pressure was significantly higher in POAG (26.1 relative unit)than that in age-matched controls (6.1 relative unit)24. The obstruc-tion might be manifested by the enlargement of the retinal veindiameter. To the contrary, BMES did not show an enlarged retinalvein or decreased AVR in POAG8. In our study, the retinal veindiameter in POAG and PACG was narrower than that in normalcontrol. The AVR in the POAG was comparable to that in normalcontrol, PACS, and PAC. Thus, BMES and HESmay not support theprevious notion that a higher retinal vein collapse pressure presentsin the glaucomatous eyes.In the HES, the mean CRAE of non-glaucomatous subjects was

156.09 mm, which is significantly narrower than that of the BMES(194.04 mm). Themultivariate-adjusted CRVE in non-glaucomawas238.25 mm, which is larger than that in the BMES (225.58 mm). TheAVR in HES (0.66) was smaller than that in BMES (0.86) or BDES(0.69)26. Our data was similar to a study conducted in Asian Malaypopulation aged 40–80 year-old27, with an AVR of 0.64 in normalcontrol and 0.65 in glaucoma. A large vein diameter can be caused bygenetic factors or an increased venous pressure; and the latter maycorrelate with the pathogenesis of PACG and other eye diseases.Whether the difference in AVR or the retinal arterial/vein bloodpressure are related to ethnicity needs further investigation.A retinal vessel with white sheath was detected in a PACG patient

in the HES. This may cause a potential difficulty in setting-up vas-cular boundaries and thus introduce inaccuracy in themeasurementsof retinal vessel diameters. Among 6716 participants, the retinalvessel was non-gradable in a PACG due to the whit sheath of theretinal vessel. To reduce the bias between different groups, graderswere trained at the Retinal Vascular Imaging Center at the Universityof Melbourne.In conclusion, using the same computer-assistant method as the

BMES and the BDES, we found that PACG and POAG patients hadsignificantly narrower retinal vessel diameters than those of the nor-mal control, PACS and PAC, after adjusting for age, gender, IOP, SE,DM, and hypertension. The AVR in our population was similar tothat of Asia Malay ethnicity but smaller than that in the Caucasians.We propose that the retinal vessel narrowing in glaucoma is second-ary to the loss of retinal ganglion cells.

1. Iwase, A. et al. The prevalence of primary open-angle glaucoma in Japanese: theTajimi Study. Ophthalmology 111, 1641–1648 (2004).

2. Wolfs, R. C. et al. Changing views on open-angle glaucoma: definitions andprevalences--The Rotterdam Study. Invest Ophthalmol Vis Sci. 41, 3309–3321(2000).

3. Wolfs, R. C. et al. Genetic risk of primary open-angle glaucoma. Population-basedfamilial aggregation study. Arch Ophthalmol. 116, 1640–1645 (1998).

4. Wilson, M. R. et al. A case-control study of risk factors in open angle glaucoma.Arch Ophthalmol. 105, 1066–1071 (1987).

5. Sommer, A. et al. Relationship between intraocular pressure and primary openangle glaucoma among white and black Americans. The Baltimore Eye Survey.Arch Ophthalmol. 109, 1090–1095 (1991).

6. Quigley, H. A. Number of people with glaucoma worldwide. Br J Ophthalmol. 80,389–393 (1996).

7. Liang, Y. B. et al. Prevalence and causes of low vision and blindness in a ruralchinese adult population: the Handan Eye Study.Ophthalmology 115, 1965–1972(2008).

8. Mitchell, P. et al. Retinal vessel diameter and open-angle glaucoma: the BlueMountains Eye Study. Ophthalmology 112, 245–50 (2005).

9. Berisha, F. et al. Retinal blood flow and nerve fiber layer measurements in early-stage open-angle glaucoma. Am J Ophthalmol. 146, 466–472 (2008).

10. Aung, T. et al. Long-term outcomes in asians after acute primary angle closure.Ophthalmology 111, 1464–9 (2004).

11. Gazzard, G. et al. Intraocular pressure and visual field loss in primary angle closureand primary open angle glaucomas. Br J Ophthalmol. 87, 720–725 (2003).

12.Quigley, H. A., Friedman, D. S. &Congdon,N. G. Possiblemechanisms of primaryangle-closure and malignant glaucoma. J Glaucoma 12, 167–80 (2003).

13. Zhou, Q. et al. The epidemiology of age-related eye diseases in Mainland China.Ophthalmic Epidemiol. 14, 399–407 (2007).

14. Liang, Y. B. et al. Rationale, design, methodology, and baseline data of apopulation-based study in rural China: the Handan Eye Study. OphthalmicEpidemiol. 16, 115–127 (2009).

15. Duan, X. R. et al. Prevalence and associations of epiretinal membranes in a ruralChinese adult population: the Handan Eye Study. Invest Ophthalmol Vis Sci 50,2018–2023 (2009).

16. Wong, T. Y. et al. Computer-assisted measurement of retinal vessel diameters inthe Beaver Dam Eye Study: methodology, correlation between eyes, and effect ofrefractive errors. Ophthalmology 111, 1183–1190 (2004).

17. Spaeth, G. L. Classification and management of patients with narrow or closedangles. Ophthalmic Surg 9, 39–44 (1978).

18. Liang, Y. B. et al. Prevalence of Primary Open Angle Glaucoma in a Rural AdultChinese Population: The Handan Eye Study. Invest Ophthalmol Vis Sci 52,8250–8257 (2010).

19. Wang, S., Xu, L., Wang, Y. & Jonas, J. B. Retinal vessel diameter in normal andglaucomatous eyes: the Beijing eye study. Clin Experiment Ophthalmol 35,800–807 (2007).

20.Wang, Y. X., Xu, L., Yang,H. & Jonas, J. B. Prevalence of glaucoma inNorthChina:the Beijing Eye Study. Am J Ophthalmol 150, 917–924 (2010).

21. Rankin, S. J. & Drance, S. M. Peripapillary focal retinal arteriolar narrowing inopen angle glaucoma. J Glaucoma 5, 22–28 (1996).

22. Jonas, J. B. & Naumann, G. O. Parapapillary retinal vessel diameter in normal andglaucoma eyes. II. Correlations. Invest Ophthalmol Vis Sci 30, 1604–1611 (1989).

23. Ikram,M. K. et al. Retinal vessel diameters and incident open-angle glaucoma andoptic disc changes: the Rotterdam study. Invest Ophthalmol Vis Sci 46, 1182–1187(2005).

24. Jonas, J. B. Central retinal artery and vein collapse pressure in eyes with chronicopen angle glaucoma. Br J Ophthalmol 87, 949–951 (2003).

25.Wittstrom, E. et al. Electrophysiological evaluation and visual outcome in patientswith central retinal vein occlusion, primary open-angle glaucoma and neovascularglaucoma. Acta Ophthalmol 88, 86–90 (2010).

26. Lee, K. E. et al. Familial aggregation of retinal vessel caliber in the beaver dam eyestudy. Invest Ophthalmol Vis Sci 45, 3929–3933 (2004).

27. Amerasinghe, N. et al. Evidence of retinal vascular narrowing in glaucomatouseyes in an Asian population. Invest Ophthalmol Vis Sci 49, 5397–5402 (2008).

AcknowledgmentsThis study was supported by National Basic Research Program of China (973 Program)Grant 2007CB512201 from the Ministry of Science &Technology of the People’s Republicof China; program of Health Policy for Blindness Prevention from Ministry of Health thePeople’s Republic of China; Key Technologies R&D Program (No. 2006–10903) fromBureau of Science and Technology of Handan city, Hebei Province, China; Beijing TongrenHospital and the Key discipline fund of Bureau of Health, Handan city, Hebei Province,China.

Author contributionsJ.G., Y.L. and N.W. wrote the main manuscript text. R.S., T.W. and F.W. analyzed fundusand retinal vessel diameter. N.W. and D.F. contributed to the diagnosis of glaucoma. Y.P.analyzed data and prepared table and figures. All authors reviewed the manuscript.

Additional informationCompeting financial interests: The authors declare no competing financial interests.

How to cite this article: Gao, J. et al. Retinal Vessels Change in Primary Angle-ClosureGlaucoma: The Handan Eye Study. Sci. Rep. 5, 9585; DOI:10.1038/srep09585 (2015).

This work is licensed under a Creative Commons Attribution 4.0 InternationalLicense. The images or other third party material in this article are included in thearticle’s Creative Commons license, unless indicated otherwise in the credit line; ifthe material is not included under the Creative Commons license, users will needto obtain permission from the license holder in order to reproduce thematerial. Toview a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

www.nature.com/scientificreports

SCIENTIFIC REPORTS | 5 : 9585 | DOI: 10.1038/srep09585 6